Downloaded from genesdev.cshlp.

org on December 15, 2016 - Published by Cold Spring Harbor Laboratory Press

Essential roles for the splicing regulator

nSR100/SRRM4 during nervous system
development
Mathieu Quesnel-Vallires,1,2,3 Manuel Irimia,2,4 Sabine P. Cordes,1,3 and Benjamin J. Blencowe1,2
1

Department of Molecular Genetics, University of Toronto, Toronto, Ontario M5S 1A8, Canada; 2Donnelly Centre, University of
Toronto, Toronto, Ontario M5S 3E1, Canada; 3Lunenfeld-Tanenbaum Research Institute, Mount Sinai Hospital, Toronto, Ontario
M5G 1X5, Canada

Alternative splicing (AS) generates vast transcriptomic complexity in the vertebrate nervous system. However, the
extent to which trans-acting splicing regulators and their target AS regulatory networks contribute to nervous
system development is not well understood. To address these questions, we generated mice lacking the vertebrateand neural-specific Ser/Arg repeat-related protein of 100 kDa (nSR100/SRRM4). Loss of nSR100 impairs
development of the central and peripheral nervous systems in part by disrupting neurite outgrowth, cortical layering
in the forebrain, and axon guidance in the corpus callosum. Accompanying these developmental defects are
widespread changes in AS that primarily result in shifts to nonneural patterns for different classes of splicing events.
The main component of the altered AS program comprises 3- to 27-nucleotide (nt) neural microexons, an emerging
class of highly conserved AS events associated with the regulation of protein interaction networks in developing
neurons and neurological disorders. Remarkably, inclusion of a 6-nt, nSR100-activated microexon in Unc13b
transcripts is sufficient to rescue a neuritogenesis defect in nSR100 mutant primary neurons. These results thus
reveal critical in vivo neurodevelopmental functions of nSR100 and further link these functions to a conserved
program of neuronal microexon splicing.
[Keywords: alternative splicing; SR proteins; nervous system development; microexons]
Supplemental material is available for this article.
Received November 18, 2014; revised version accepted February 27, 2015.

Proper development and function of the mammalian nervous system depend on the tight coordination of multiple
layers of gene regulation. During development, neurons
progress through maturation stages to acquire their subtype-specific functions (Gao et al. 2013). For example,
neurons born in the subventricular zone (SVZ) of the embryonic cortex use endocrine and exocrine cues while migrating dorsally to establish and populate specific cortical
layers. Similarly, neuronal projections in the brain and periphery rely on successive adjustments of intrinsic to extrinsic factors to arrive at their targets. Considerable
remodeling of the cytoskeleton, vesicular transport, and
other subcellular processes allows neurons to achieve
their designated morphologies and functions. While similar repertoires of genes are associated with these processes, it is becoming apparent that extensive variation at
the level of post-transcriptional regulation generates the
remarkable transcriptomic and proteomic diversity required for establishing biological complexity during verte4

Present address: EMBL/CRG Research Unit in Systems Biology, Centre

for Genomic Regulation (CRG), Barcelona 08003, Spain
Corresponding authors: b.blencowe@utoronto.ca, cordes@lunenfeld.ca
Article is online at http://www.genesdev.org/cgi/doi/10.1101/gad.256115.
114.

746

brate nervous system development (Li et al. 2007; Norris

and Calarco 2012; Lipscombe et al. 2013a; Zheng and
Black 2013).
Alternative splicing (AS) is the process by which different pairs of splice sites are selected to produce multiple
transcripts from a single gene. It is controlled by the concerted action of multiple cis-acting motifs and cognate
trans-acting factors that promote or repress the assembly
of productive splicing complexes (spliceosomes) at splice
sites (Chen and Manley 2009; Braunschweig et al. 2013).
Widely expressed and tissue-restricted RNA-binding proteins combine to regulate AS decisions via positive- and
negative-acting cis motifs located in exons or flanking introns, referred to as splicing enhancers and silencers, respectively. AS represents a major source of biological
diversity that likely afforded the evolution of complexity
associated with the development and function of the
vertebrate nervous system (Barbosa-Morais et al. 2012;
2015 Quesnel-Vallires et al. This article is distributed exclusively by
Cold Spring Harbor Laboratory Press for the first six months after the fullissue publication date (see http://genesdev.cshlp.org/site/misc/terms.
xhtml). After six months, it is available under a Creative Commons License (Attribution-NonCommercial 4.0 International), as described at
http://creativecommons.org/licenses/by-nc/4.0/.

GENES & DEVELOPMENT 29:746759 Published by Cold Spring Harbor Laboratory Press; ISSN 0890-9369/15; www.genesdev.org

Downloaded from genesdev.cshlp.org on December 15, 2016 - Published by Cold Spring Harbor Laboratory Press

In vivo functions of nSR100

Merkin et al. 2012). Indeed, AS patterns are more complex

in the brain than other tissues, and many of these events
happen in genes implicated in complex neuronal processes, such as the control of synaptic plasticity associated
with cognition (Lipscombe 2005; Ule and Darnell 2006).
While most tissue differential splicing patterns are species-specific in vertebrates, there is a higher frequency of
conserved alternative cassette exon inclusion events in
vertebrate brains than in other tissue types (Barbosa-Morais et al. 2012; Merkin et al. 2012). This suggests the existence of a core set of conserved functions for AS across
vertebrate species in addition to roles for AS underlying
species-specific neurodevelopmental and behavioral characteristics. However, little is known about the in vivo
functions of the protein factors that are responsible for establishing AS complexity in the nervous system or the
functions of the individual AS events that are controlled
by these factors.
Neural-enriched splicing regulators, including the
Nova, Rbfox, and Ptbp proteins, have been characterized
using mouse models. Nova proteins, which were originally identified as the autoantigens in patients with paraneoplastic opsoclonus myoclonus ataxia (Buckanovich et al.
1993; Yang et al. 1998), control the inhibitory synapse, and
their knockout results in cortical migration (Yano et al.
2010) and neuromuscular junction (NMJ) defects (Ruggiu
et al. 2009). Rbfox1 and Rbfox2 mutant mice are susceptible to seizures and display disrupted cerebellar development (Gehman et al. 2012). Depending on the strain
background, Ptbp2 knockout mice die at birth or else exhibit cortical degeneration and lethality during the first
few postnatal weeks (Licatalosi et al. 2012; Li et al.
2014). Additional studies using Nova knockout mice
have revealed functions for specific Nova-regulated splice
variants (including alternative exons in the Dab1 gene)
that facilitate the proper migration of newly born cortical
neurons (Yano et al. 2010) and exons in the Agrin gene that
are important for the formation of NMJs (Ruggiu et al.
2009). However, aside from these examples, few other
neuronal genes have been characterized at isoform resolution in vivo (Norris and Calarco 2012; Lipscombe et al.
2013b; Zheng and Black 2013).
We previously identified and characterized the vertebrate- and neural-specific Ser/Arg repeat-related protein
of 100 kDa (nSR100/SRRM4) (Calarco et al. 2009; Raj
et al. 2011, 2014). Knockdown and overexpression experiments performed in cell culture revealed that nSR100 promotes the inclusion of 30%50% of the conserved human
and mouse cassette alternative exons that display brainspecific inclusion patterns in transcriptome profiling
data (Raj et al. 2014). Knockdown of nSR100 in Neuro2a
cells and developing zebrafish was shown to impair neurite outgrowth and branching of trigeminal ganglia, respectively (Calarco et al. 2009), and in utero knockdown
of nSR100 in mice prevented differentiation of neuronal
progenitors in the cortex (Raj et al. 2011). Recently, the
Bronx waltzer (bv) mouse mutation was mapped to
the nSR100 gene (Nakano et al. 2012). bv homozygotes
display hearing and balance defects attributed to degeneration of inner ear hair cells. The apparent limited pheno-

typic consequences of the bv mutation are likely

because this mutation eliminates only the terminal
exon and part of the 3 untranslated region (UTR) of
nSR100 transcripts, leaving the majority of the nSR100
protein intact.
nSR100-regulated exons were found to be concentrated
in genes that function in various aspects of neuronal development and function (Calarco et al. 2009; Raj et al.
2011, 2014; Nakano et al. 2012). These and other neuralregulated exons that are >27 nucleotides (nt) in length
are highly concentrated in surface-accessible disordered
regions of proteins and function in the regulation of proteinprotein interactions (Buljan et al. 2012; Ellis et al.
2012). Furthermore, in a very recent study, we showed
that nSR100 strongly promotes the inclusion of very
short, 3- to 27-nt, neuronal microexons (Irimia et al.
2014). The corresponding microexon residues are concentrated withinor immediately adjacent toproteinprotein or proteinlipid interaction domains. Most of these
exons display striking increases in inclusion during neuronal maturation, coincident with increased expression of
nSR100. Notably, they also show significant decreases
in inclusioncoincident with reduced expression of
nSR100in the cortices of individuals with autism spectrum disorder (ASD) (Irimia et al. 2014). A key function of
nSR100 thus appears to be the widespread regulation of
protein interactions required for the maturation and proper function of neurons. However, the scope of the in vivo
functions of nSR100 during nervous system development
has not been previously addressed.
To investigate the functions of nSR100 in vivo, we generated mice carrying a conditional exon deletion in the
nSR100 (Srrm4) gene that results in widespread loss of
the full-length protein. We observed that nSR100 is essential for early postnatal survival of a large majority of mutant animals, with the few surviving animals displaying
balance defects similar to those seen in bv/bv mice but
also exhibiting persistent tremors. Additionally, loss of
nSR100 in mice results in impaired neurite outgrowth in
the diaphragm, early neuronal commitment of neural progenitors leading to defective cortical layering, and a failure
of callosal axons to cross the midline in the forebrain. Using RNA sequencing (RNA-seq) profiling, we defined all
classes of AS (including alternative microexons and longer
alternative cassette exons, 5 and 3 splice sites, and retained introns) that are controlled by nSR100 in vivo, a
great majority of which were not previously reported. A
large fraction of alternative cassette exons and microexons positively regulated by nSR100 are neurally enriched,
which is not the case for other classes of nSR100-dependent splicing events. Moreover, a higher proportion of
neural microexons is affected by disruption of nSR100
than are other neural-regulated AS events. These include
highly conserved exons with the potential to insert only
one or two amino acids in proteins of key functional relevance to neuronal maturation. Remarkably, an nSR100regulated 6-nt microexon in the Unc13b gene promotes
neurite growth in mouse primary neurons. Cortical neurons from nSR100 78/78 mice display a neuritogenesis
defect, and expression of Unc13b transcripts including

GENES & DEVELOPMENT

747

Downloaded from genesdev.cshlp.org on December 15, 2016 - Published by Cold Spring Harbor Laboratory Press

Quesnel-Vallires et al.

the microexon, but not transcripts lacking the microexon,

is sufficient to rescue the mutant phenotype. Collectively,
our results define critical new in vivo functions of nSR100
during mouse development and further link these functions to the disruption of a conserved program of
nSR100-dependent neuronal microexons.
Results
Perinatal mortality in nSR100 mutant mice
Our previous studies using in vivo knockdown of nSR100
in zebrafish and mouse embryos suggested that nSR100
may play a role in several aspects of nervous system development (Calarco et al. 2009; Raj et al. 2011). To address
the extent of nSR100 functions in the developing nervous
system, we generated mice carrying a conditional knockout nSR100 lox allele from embryonic stem cells obtained
from EUCOMM (European Conditional Mouse Mutagenesis Program). The nSR100 lox allele includes a gene trap
upstream of a LacZ reporter and LoxP sites framing
nSR100 exons 7 and 8 (Fig. 1A). Southern blotting confirmed the integrity of the integration site in nSR100 lox
mice (Fig. 1B). We crossed nSR100 +/lox mice with mice
carrying the widely expressed CMV-Cre recombinase
transgene to obtain nSR100 78 mice, in which exons 7
and 8 have been deleted throughout the animal and in
the germline. This deletion introduces a +2 frameshift in
downstream exons and causes complete loss of full-length
nSR100 transcripts and protein in homozygous nSR10078

mice (Fig. 1C,D). Immunoblotting revealed that a 25-kDa

protein fragment could be detected in homozygous and
heterozygous mutant mice using an antibody to the N terminus of nSR100. RTPCR confirmed that transcripts encompassing nSR100 exons 16 and the gene trap insertion
are preserved in the mutant mouse (Supplemental Fig.
S1A). The small N-terminal fragment produced from the
nSR100 78 allele lacks the RS-rich domain of nSR100
(Supplemental Fig. S1B), which, based on previous studies
of nSR100 and other SR proteins, is predicted to function
in the formation of proteinprotein and/or proteinRNA
interactions required for splicing complex formation
(Wu and Maniatis 1993; Shen and Green 2004; Raj et al.
2014). In contrast to the full-length protein, overexpression of the truncated protein in Neuro2a cells depleted
of endogenous nSR100 fails to restore nSR100-dependent
splicing (Supplemental Fig. S1C, lanes 13). Moreover,
when coexpressed with full-length nSR100 in Neuro2a
cells, the truncated mutant does not interfere with splicing of nSR100 target exons (Supplemental Fig. S1C, lanes
46). These observations provide evidence that the
25-kDa protein fragment lacks critical functional activities associated with full-length nSR100.
We observed that >85% of nSR100 78/78 mice died
in the first few hours after birth (Supplemental Table
S1). Although these mice present no gross morphological
phenotype at late embryonic stages or at birth (Fig. 1E),
they show signs of respiratory defects, including irregular
breathing and heavy gasping, and become cyanotic soon
after birth. This phenotype contrasts sharply with those

Figure 1. Loss of the full-length nSR100

protein in nSR100 78/78 mutant mice.
(A, top panel) Map of the conditional
nSR100/SRRM4 allele showing the positions of exons, Frt (open triangles) and
LoxP (solid triangles) recombination sites,
homology arms (dashed boxes), and cutting
sites for the AseI restriction enzyme (vertical arrows) and the probe (solid bar) used
for Southern blot analysis (see B). (Bottom
panel) Map of the knockout allele following
crossing of the conditional nSR100 lox
mouse with a CMV-Cre transgenic line.
Cre-LoxP recombination drives the loss of
nSR100 exons 7 and 8 and results in a +2
frameshift and the introduction of several
premature termination codons downstream
from the deletion. The position of primers
used for RTPCR (horizontal arrows; see
C) is indicated. Homozygous nSR100 lox/lox
mice do not display any overt phenotype.
(B) Southern blot analysis on tail DNA
from wild-type (+/+), conditional (lox), and
knockout (78) mice. DNA was digested
with AseI and hybridized with a probe binding upstream of the 5 homology arm on the
conditional allele in intron 3. Predicted band size is 15.4 kb in wild-type, 16.4 kb in conditional, and 19.4 kb in knockout alleles, respectively. (C ) RTPCR on embryonic day 16.5 (E16.5) whole-brain total RNA using primers amplifying exon 2 to exon 9. No transcript could
be detected in homozygous mutants. (D) Immunoblotting on E17.5, whole-brain lysates using an antibody to nSR100. Full-length nSR100
protein is completely lost in homozygous mutants (arrow), but a 25-kDa fragment is expressed from the 78 allele. (E) E17.5 mutant embryos display normal morphology.

748

GENES & DEVELOPMENT

Downloaded from genesdev.cshlp.org on December 15, 2016 - Published by Cold Spring Harbor Laboratory Press

In vivo functions of nSR100

of the previously described nSR100 mutant bv mouse, in

which only the last 103 amino acids from the C terminus
of nSR100 are lost. Homozygous bv mice are viable and
display a phenotype limited to the degeneration of the inner hair cells of the inner ear (Deol and GluecksohnWaelsch 1979; Nakano et al. 2012). Notably, the few homozygous nSR100 78/78 survivors that we obtained
from crossing heterozygous parents display a head tilting
and circling behavior reminiscent of the balancing defect
observed for the bv mutant strain. However, in contrast to
the bv mutant, all surviving nSR100 78/78 individuals
additionally display pronounced tremors, a phenotype
that is often associated with neurobiological defects (Supplemental Videos S1S3). Homozygous mutants escaping
perinatal mortality did not display significant differences
in life span compared with wild-type littermates (data not
shown). Embryos harvested at embryonic day 17.5 (E17.5)
and E18.5 were found at Mendelian ratios (Supplemental
Table S1), indicating that loss of nSR100 does not cause
early embryonic lethality. The extensive perinatal mortality and neurobiological phenotypes observed in surviving
nSR100 78/78 mice highlight the importance of nSR100
during embryonic development as well as its role in the
proper functioning of the adult nervous system.
Loss of nSR100 impairs diverse neuronal processes
The respiratory problems accompanying perinatal mortality in nSR100 78/78 mice suggested that the innervation
of the diaphragm might be impaired by loss of the nSR100
protein. We first asked whether nSR100 is expressed in the
peripheral nervous system where motor neurons innervating the diaphragm are located. We surveyed nSR100 expression at different time points during development
using both the LacZ cassette in the nSR100 lox mouse as
a reporter for nSR100 gene expression and in situ RNA
hybridization in wild-type mice. X-Gal staining and in
situ hybridization showed that nSR100 is expressed in

both the brain and the neural tube during early neurogenesis, with nSR100 mRNA expressed as early as E8.5 and
LacZ reporter expression starting at E9.5 (Supplemental Figs. S2A,B, S3A). Immunostaining of sections from
nSR100 +/lox mice using anti--galactosidase antibody and
anti-NeuN antibody to mark post-mitotic neurons revealed that only neurons express nSR100 in the brain
(Supplemental Fig. S3B). Moreover, in situ hybridization
at E17.5 showed that nSR100 expression is maintained
in the brain during development, with high expression
in the cerebral cortex and hippocampus late in embryogenesis (Supplemental Fig. S2C). These results are consistent with recent analyses of RNA-seq data from diverse
human and mouse cell and tissue types and a neuronal differentiation time series, which indicated that nSR100 expression is neuron-specific, occurs in the brain and dorsal
root ganglia, and increases in the brain from E11 to E18 before decreasing in the adult (Irimia et al. 2014; Raj et al.
2014). Taken together, these data confirm that nSR100
is neuronal-specific and is expressed in both the central
and peripheral nervous system in developing mice.
We next visualized the innervation of the diaphragm
just before birth at E18.5 using an antibody to neurofilament on whole-mount preparations. This staining revealed that primary branches deriving from the phrenic
nerve appear thinner in nSR100 78/78 mice (Fig. 2A).
In addition, we observed that the total length covered by
secondary motor axons is greatly reduced and that the
number of secondary axons is decreased by almost twofold
in homozygous mutants, a phenotype not seen in heterozygotes (Fig. 2B,C). These defects are already present at
E16.5 (Supplemental Fig. S4AC), suggesting that the
lack of secondary branches does not stem from degeneration or pruning but rather from deficient sprouting in the
mutant mice. The overall distance covered by primary axons was not affected at either E18.5 or E16.5 (Fig. 2D; Supplemental Fig. S4D). Each individual secondary branch
forming in the mutants projects as far as its wild-type

Figure 2. Loss of nSR100 impairs neurite

outgrowth in motor neurons. (A) Wholemount staining of E18.5 diaphragms with
anti-neurofilament antibody (green) to
highlight innervation. Orange dots mark
secondary branches in the insets. Bars: left
panels, 1000 m; insets, 500 m. (B,C )
The total distance covered by all secondary
axons (B) and the number of secondary
branches present on the right ventral primary branch of the phrenic nerve (C ) were
quantified on three or four individuals for
each genotype. The total distance covered
by secondary neurites and the number of
secondary branches formed are significantly lower in homozygous mutants. (D)
The total length covered by primary
branches is not affected in homozygous
mutants. (E) The average length of individual secondary branches in the mutant does not differ significantly from those of wild-type and heterozygous littermates. Three diaphragms from wild-type and heterozygous embryos and four diaphragms for homozygous mutants were analyzed. One-way ANOVA
with Tukey-Kramer post-hoc test. Error bars indicate standard error of the mean.

GENES & DEVELOPMENT

749

Downloaded from genesdev.cshlp.org on December 15, 2016 - Published by Cold Spring Harbor Laboratory Press

Quesnel-Vallires et al.

counterpart (Fig. 2E; Supplemental Fig. S4E), and motor

endplates form in the same numbers in the diaphragm of
nSR100 homozygous and heterozygous mutant mice, although at higher density in the homozygous mutant,
most likely due to a lack of secondary branching (Supplemental Fig. S4FH). The diminished axon sprouting capacities of motor neurons in the diaphragm of nSR100 78/78
mice likely contributes to nSR100-dependent respiratory
defects and early postnatal death. These axon guiding or
branching defects are not limited to phrenic nerve innervation, as we detected defective formation of the trigeminal,
hypoglossal, and spinal nerves in whole-mount staining of
E10.5 and E12.5 embryos (Supplemental Fig. S5).
Because nSR100 is highly expressed in the cortex and in
utero knockdown of nSR100 resulted in defects in neuronal differentiation in the cortex (Raj et al. 2011), we investigated whether cortical anatomy was altered in nSR100
mutants. The establishment of defined cortical layers is
an important and conserved step in mammalian brain

development. Immunofluorescence using layer-specific

markers revealed that the deep, T-box brain 1 (Tbr1)-positive cortical layer VI is enlarged and comprised of more
cells in the homozygous mutant, a phenotype also seen
to a lesser extent in heterozygotes (P < 0.0001, one-way
ANOVA) (Fig. 3A,B). The definition of the preplate was
also altered in homozygotes and heterozygotes. Costaining with antibodies to Tbr1 and Satb2 to highlight superficial layers IIV revealed a decrease in the number of
superficial neurons in nSR100 mutants (P < 0.05, oneway ANOVA) (Fig. 3A,C). The total number of post-mitotic neurons was also reduced in the mutants, as highlighted by NeuN staining (P < 0.0001, one-way ANOVA)
(Fig. 3A,D).
We postulated that the increase in the number of earlyborn neurons and decrease in the number of late-born
neurons observed in nSR100 mutants may be due to
the premature commitment of neural progenitors to
a post-mitotic fate. Indeed, Pax6 staining at E18.5

Figure 3. nSR100 mutant mice display

aberrant cortical layering and premature
neurogenesis. (A) Immunofluorescence microscopy using antibodies to Tbr1, Satb2,
NeuN, and Pax6 to label deep layer VI,
superficial layers IIV, post-mitotic neurons, and neural progenitors, respectively,
on coronal sections of E18.5 embryonic
brains. Bars, 50 m. (IVI) Cortical layers
IVI, (CM) cortical mantle. Dashed white
lines highlight ventral and dorsal cortical
boundaries. (BE) The number of Tbr1+ (B),
layer IIV Satb2+ (C ), NeuN+ (D), and
Pax6+ (E) cells was quantified for three to
five individuals per genotype and on three
sections for each individual. These immunostaining experiments highlight an increase in the number of deep, early-born
Tbr1+ neurons and a corresponding decrease
in superficial Satb2+ neurons as well as a
decrease in the total number of neurons
(NeuN+) and neural progenitors (Pax6+). (F)
EdU-labeling was performed at E12.5
(green), and brains were harvested at E18.5
and stained with an antibody to Tbr1 (red).
Bar, 100 m. (GI ) The number of EdU+ cells
was counted in deep layer VI (G), superficial
layers IIV (H), and the SVZ (I). (J) The thickness of the SVZ was measured from the preplate to the lateral ventricle and relative to
the total thickness of the cortex measured
from the surface of layer I to the lateral ventricle. Four to five embryos per genotype
and three sections per embryo were analyzed. Whiskers indicate the 10th and 90th
percentiles in all box plots. One-way
ANOVA with Tukey-Kramer post-hoc test.

750

GENES & DEVELOPMENT

Downloaded from genesdev.cshlp.org on December 15, 2016 - Published by Cold Spring Harbor Laboratory Press

In vivo functions of nSR100

confirmed that the pool of neural progenitors is depleted

in nSR100 78/78 mice at the end of embryogenesis
(P < 0.05, one-way ANOVA) (Fig. 3A,E). To verify whether
deregulation of neurogenesis was the major cause for the
cortical layering phenotype, we performed 5-ethynyl-2 deoxyuridine (EdU) labeling at E12.5, when early neurons
populating deep cortical layers are born, and harvested the
labeled brains at E18.5. Costaining for EdU+ and Tbr1+
cells confirmed that a greater number of Tbr1+ cells are
born at E12.5 in nSR100 mutants (P < 0.05, one-way
ANOVA) (Fig. 3F,G). The increase in the number of
EdU+ cells was specific to the deep cortical layer VI, as a
similar number of EdU+ cells was observed in the SVZ
and superficial cortical layers across all genotypes (Fig.
3H,I). Finally, we detected an overall thinning of the
SVZ in nSR100 mutant brains (P < 0.05, one-way ANOVA)
(Fig. 3J) that parallels the early post-mitotic commitment
of progenitors and the reduction of their numbers. Taken
together, these results indicate that the cortical layering
defects in nSR100 78 mice are primarily caused by deregulation of the timing of early neurogenesis.
Interestingly, while analyzing cortical layering, we noticed that the morphology of the rostral part of the corpus
callosum in nSR100 78/78 mice differed from its stereotypical shape. The corpus callosum consists mostly of cortical axons crossing the midline to contact neurons of the
opposite hemisphere. This interconnection between
hemispheres is essential for the fast processing of information and cognition (Paul et al. 2007). Neurofilament
immunostaining revealed that several callosal axons are
misguided in the absence of nSR100 and form thick ectopic fascicles similar to Probst bundles, projecting ventrally
instead of crossing the midline (Fig. 4A). This phenotype is
never observed in wild-type mice but is important enough
in homozygous mutants to alter the shape of the corpus
callosum. Although the corpus callosum of nSR100 +/78
mice does not appear grossly misshapen, it also contains
ectopic ventrally projecting bundles (Fig. 4B). These observations represent the first example of a midline crossing defect as a consequence of the knockout of an AS
regulator. Overall, our phenotypic survey so far shows
that nSR100 controls a diverse array of neuronal functions
in both the central and peripheral nervous systems, including cortical layering, axon guidance, and midline
crossing.

An in vivo nSR100-regulated splicing program

To identify AS events that contribute to the aforementioned neurodevelopmental deficits, we performed
RNA-seq analyses on two sets of biological replicate samples, each consisting of pooled E18.5 mouse cortical or
hippocampal tissue from wild-type and nSR100 78/78
mice (eight samples in total). An RNA-seq analysis pipeline was employed that generates quantitative estimates
for percent spliced in (PSI) values for alternative cassette exons, percent splice site usage (PSU) values for
sequences formed by alternative 5 /3 splice site selection,
and percent intron retained (PIR) values for intron retention events (Braunschweig et al. 2014; Irimia et al.
2014). This pipeline also identifies and quantifies PSI values for 3- to 27-nt microexons (Irimia et al. 2014). To identify which AS events were differentially spliced between
wild-type and nSR100 78/78 brains, we performed a
paired t-test between the four pairs of samples and required an average PSI/PIR/PSU between pairs of samples
of 10. Of the 263 AS events displaying differential splicing according to these criteria (Supplemental Table S2),
cassette alternative exons, including microexons, represented the largest class, comprising 58% of the total
(Fig. 5A). A large number of retained introns as well as a
few alternative 3 and 5 alternative splice sites were also
misregulated in nSR100 78/78 brain tissues. Of the alternative cassette exons and microexons that displayed
changes, 66 (83%) and 72 (100%), respectively, displayed
decreased PSI levels in nSR100 78/78 mouse brains. Furthermore, 70% of cassette and 96% of microexons positively regulated by nSR100 were defined by RNA-seq
profiling of multiple wild-type mouse tissues (Irimia
et al. 2014) as having increased neural inclusion compared
with other tissues (Fig. 5B). Other classes of AS events displaying differential splicing in nSR100 78/78 mice did
not display enrichment for neural-specific regulation or
nSR100-dependent inclusion.
An analysis of the cumulative distributions of exon
lengths for cassette exons shows that nSR100-regulated
exons are significantly shorter than the full set of neurally
regulated exons with either increased or decreased neural
PSI as well as nonneural alternative exons (P < 107 for all
comparisons with nSR100-regulated exons; Wilcoxon
rank-sum tests) (Fig. 5C). Moreover, consistent with our

Figure 4. Midline crossing defects in

nSR100 mutant mice. (A) Negative gray70
scale images of immunofluorescence mi60
croscopy
using
an
antibody
to
50
40
neurofilament on coronal sections of the
30
rostral part of the corpus callosum of
20
10
E18.5 embryos. Dashed lines with arrow0
heads show either the prototypical tracts
-8
-8
7
7
/
+/
-8
of callosal axons in the wild-type (+/+) or
7

the ectopic ventral projections in the homozygous mutant (78/78). Arrows indicate ectopic bundles in the heterozygous and homozygous mutants. Bar, 100 m. (B) The thickness of ventrally projecting bundles was measured at three levels on each side of the corpus
callosum for three (+/78) or four (78/78) individuals per genotype and on three sections for each individual. Whiskers indicate
the 10th and 90th percentiles. One-tailed Mann-Whitney test.
+/+

+/ 7-8

7-8/ 7-8

80

p < 0.0001

Thickness of
bundles (m)

GENES & DEVELOPMENT

751

Downloaded from genesdev.cshlp.org on December 15, 2016 - Published by Cold Spring Harbor Laboratory Press

Quesnel-Vallires et al.

Figure 5. nSR100 regulatory program in

the mouse brain. (A) Number of AS events
showing significantly decreased (left) or increased (right) inclusion upon nSR100 deletion in mouse brains, plotted by class.
(AltEx) Alternative cassette exons. (B)
Microexons (blue) and longer cassette exons
(red) were plotted based on their PSI difference between nSR100 78/78 and wildtype samples (X-axis) and their PSI between the average of neural versus nonneural tissues (Y-axis) as previously determined
(Irimia et al. 2014). (C) Cumulative distribution of exon lengths for different groups of
alternative exons, including events that
show decreased inclusion in nSR100 78/
78
compared with the control (nSR100-enhanced; dark blue), all alternative exons
with increased neural PSI (neural increased;
light blue), all alternative exons with decreased neural PSI (neural decreased; red),
and nonneural alternative exons (nonneural; gray). (D) Cumulative distribution plots
indicating the position of the first UGC motif within 200 nt upstream of nSR100-regulated microexons (dark blue), longer exons
(>27 nt) with increased neural inclusion
(light blue), exons with decreased neural inclusion (red), and nonneural (gray) and constitutive (black) exons. The number of
exons used in the analysis for each subgroup
is indicated in parentheses. (E) RTPCR validations of nSR100-regulated cassette exons
in cortical (left two lanes) and hippocampal
(right two lanes) samples. PSI values calculated from semiquantitative RTPCR or
RNA-seq analysis are shown below the gel
for each event. Cassette exon included isoforms are represented by yellow boxes
flanked by blue constitutive exons. Primers
were located in flanking constitutive exons.

recent results from analyzing nSR100-dependent, neuralregulated exons in cell lines (Raj et al. 2014), nSR100-regulated microexons show very strong enrichment for UGC
motifs in the first several nucleotides upstream of microexons regulated by nSR100 in vivo (Fig. 5D). Of 22 analyzed differential splicing events involving alternative
cassette and microexons, which were detected by RNAseq to undergo reduced inclusion as a consequence of
the loss of nSR100, all were validated by semiquantitative
RTPCR assays, including 3- and 6-nt microexons (Fig. 5E;
Supplemental Fig. S6). Expression analysis based on
cRPKMs (corrected reads per kilobase per million mapped
reads) revealed only nine genes, other than nSR100, with
an average mRNA expression difference of 1.5-fold between both replicates of wild-type and nSR100 78/78
tissues and P < 0.05 (paired t-test) (Supplemental Table
S3). Analysis of genes with alternative cassette exons
and microexons affected by loss of nSR100 revealed significant enrichment (P < 0.01) for gene ontology (GO) terms
essential to many aspects of neuronal cell biology, such

752

GENES & DEVELOPMENT

as vesicle-mediated transport, neurotransmitter secretion, synaptosome, and cell projection morphogenesis (Supplemental Fig. S7). Collectively, these
observations suggest that multiple neural cassette exons
in particular, highly conserved microexons that display
marked decreases in inclusion levels as a consequence of
the loss of nSR100may underlie mutant phenotypes detected in nSR10078/78 mice.
Functions of nSR100-regulated microexons
Based on our previous and present analyses of the in vivo
mutant phenotypes of zebrafish and mice lacking nSR100
and also the known functions of genes that harbor
nSR100-dependent exons, a major function of the
nSR100-regulated splicing program is likely to control different aspects of neurite outgrowth. Consistent with this
proposal, we found that hippocampal neurons cultured
from nSR100 78/78 mice have significantly shorter
neurites compared with neurons from wild-type animals

Downloaded from genesdev.cshlp.org on December 15, 2016 - Published by Cold Spring Harbor Laboratory Press

In vivo functions of nSR100

(P < 0.0001, two-tailed MannWhitney test) (Fig. 6A,B). To

investigate whether nSR100-regulated microexons may
be responsible for neurite outgrowth, we focused on a previously uncharacterized, highly conserved nSR100 target
microexon of 6 nt in Unc13b/Munc13 (Fig. 5E), a gene
that has previously been shown to contribute to early neuritogenesis in primary mouse neurons (Broeke et al. 2010).
Since our RNA-seq analysis can only locate this microexon in the context of its immediate flanking constitutive
exons due to short read length, Sanger sequencing of RT
PCR products from mouse brains was performed. This revealed that the Unc13b microexon, located between exons 13 and 14, is spliced in transcripts that contain at
least exons 514 and exons 1120.
To address whether increased skipping of the Unc13b
microexon may contribute to the neuritogenesis defect
in nSR100 78/78 neurons, we harvested cortical neurons
from wild-type and mutant E18.5 embryos, transfected
them with red fluorescence protein (RFP)-Unc13b expression constructs that either include (Unc13b-inc) or skip
(Unc13b-skp) the microexon (Fig. 6C,D), and plated

them at low density. At day 2 in vitro (DIV2), the cellular

distribution and expression levels of Unc13b-inc-RFP and
Unc13b-skp-RFP appear similar (Fig. 6D,E; Supplemental
Fig. S8A). Both Unc13b isoforms are distributed over the
entire length of the axons, thereby facilitating the visualization of transfected neurons and quantification of individual neurites. Control RFP-expressing mutant cortical
neurons display the same neuritogenesis defect as hippocampal neurons (P < 0.001, Kruskal-Wallis test) (Fig. 6F).
Remarkably, wild-type neurons expressing Unc13b-skpRFP produce neurites that are as short as mutant neurons
expressing control RFP and significantly shorter than control wild-type neurons (P < 0.05, Kruskal-Wallis test) (Fig.
6E,F). Expression of the skipped Unc13b isoform does
not further affect neurite growth in mutant neurons.
Strikingly, however, while expression of Unc13-inc-RFP
does not affect neurite growth in wild-type neurons, inclusion of the microexon in mutant cells restores the lengths
of neurites to levels observed in wild-type neurons (Fig.
6E,F). Transfection of an nSR100-RFP expression vector
in mutant neurons also results in longer neurites than in

Figure 6. Functional regulation of a neural

microexon by nSR100. (A) Representative
images of primary hippocampal neurons
from wild-type (+/+) and nSR100 78/78
mice cultured for 2 d and then stained
with antibodies to Tuj1 (red) and Map2
(green). Bar, 25 m. (B) The length of the
longest neurite was measured for each neuron. Four-hundred-fifty-one cells from wildtype embryos and 425 cells from mutant
embryos were analyzed. (C) Immunoblotting with an antibody to red fluorescence
protein (RFP) on Neuro2A lysates transfected with increasing amounts of the same
constructs that were used for the experiments in DF showing Unc13b-skp-RFP
(skp) and Unc13b-inc-RFP (inc) protein expression. (D) RTPCR showing inclusion
levels of the Unc13b microexon as well as
RFP, nSR100, and GAPDH expression in
transfected nSR100 +/+ and nSR100 78/78
cortical neuronal cultures. (E) Representative images of primary cortical neurons
from nSR100 +/+ and nSR100 78/78 mice
transfected with RFP, Unc13b-skp-RFP
(skp), Unc13b-inc-RFP (inc), or nSR100RFP; cultured for 2 d; and then stained
with an antibody to Tuj1 (green). nSR100RFP displays nuclear localization, as expected. Bar, 25 m. (F ) nSR100 +/+ primary
cortical neurons were transfected with
RFP, Unc13b-skp-RFP, or Unc13b-inc-RFP
(three left groups), and nSR100 78/78 primary cortical neurons were transfected
with the same constructs and nSR100RFP (four right groups). The longest neurites were measured in RFP-expressing
cells. Whiskers indicate the 10th and 90th
percentiles. Kruskal-Wallis test with
Dunns multiple comparison tests.

GENES & DEVELOPMENT

753

Downloaded from genesdev.cshlp.org on December 15, 2016 - Published by Cold Spring Harbor Laboratory Press

Quesnel-Vallires et al.

mutant cells expressing Unc13b-skp-RFP (P < 0.05, Kruskal-Wallis test) (Fig. 6E,F). Furthermore, it is noteworthy
that untransfected wild-type neurons cultured in parallel
with neurons transfected with any construct produced
neurites of similar length, whereas shorter neurites were
consistently detected in untransfected nSR100 78/78
neurons (P < 0.0001, one-way ANOVA) (Supplemental
Fig. S8B). This confirms that the differences in neurite
length observed here are dependent on the specific construct that is expressed and are not a consequence of variability between cultures. These results thus reveal that
the inclusion of a single nSR100-dependent microexon
of 6 nt can positively stimulate neurite formation and rescue a neurite growth defect in primary nSR100 78/78
neurons. The phenotypes that we observed in nSR10078
/78
mice may therefore be attributed, at least in part, to
the reduced inclusion of neuronal microexons.

Discussion
In this study, we generated and characterized mice deficient of nSR100/SRRM4, a vertebrate- and neural-specific
splicing factor that regulates 30% of alternative exons
with increased neural inclusion, including a large number
of highly conserved 3- to 27-nt microexons. We showed
that the loss of nSR100 protein in vivo results in numerous neurodevelopmental defects during mouse embryogenesis that lead to early postnatal mortality in the
majority of animals. We further linked these neurodevelopmental deficiencies to the loss of microexon
regulation.

nSR100 regulates multiple neurodevelopmental

processes
Some of the neurodevelopmental phenotypes observed in
mice deficient of nSR100 may relate to altered phenotypes
seen in other splicing factor knockouts, while others are
unique. Neonatal lethality has been reported as a consequence of loss of the splicing regulator Ptbp2. Ptbp2 is expressed in neurons as well as in skeletal and cardiac
muscle (Licatalosi et al. 2012), and mice lacking Ptbp2
are paralyzed at birth (Licatalosi et al. 2012; Li et al.
2014). However, mice lacking Ptbp2 specifically in neurons (Ptbp2 Nestin knockout mice) die within an hour of
birth, similar to nSR100 mutants, and initiate breathing
at a greatly reduced rate (Li et al. 2014). Given that
nSR100 promotes the expression of Ptbp2 by activating
the inclusion of an alternative exon that prevents nonsense-mediated decay of Ptbp2 transcripts (see above; Calarco et al. 2009), it is possible that the requirement for
nSR100 for innervation of the diaphragm may relate, at
least in part, to Ptbp2 misregulation. However, we found
that only 32.8% of in vivo nSR100-regulated exons overlap with in vivo Ptbp2 targets. Therefore, overlapping
and distinct exons targeted by these factors may contribute to breathing defects, paralysis, and early postnatal
death. This conclusion is further supported by the obser-

754

GENES & DEVELOPMENT

vation of phenotypes that are unique to nSR100 78/78

mice.
In addition, as in the case of nSR100 78/78 and Ptbp2
knockout mice, mice deficient in both Nova1 and Nova2
proteins (Nova double-knockout mice) showed respiratory defects at birth (Ruggiu et al. 2009). While phrenic
nerve branching appeared normal, NMJs in E18.5 Nova
double-knockout mice had few acetylcholine receptors
(AChRs) and only rarely contacted motor axon terminals.
In contrast, in nSR100 78/78 mice, motor endplates (sarcolemma folds in which AChRs concentrate) are of an
abundance similar to those of nSR100 +/78 heterozygous
littermates, which are fully viable. However, the distribution of AChRs is altered due to the phrenic nerve deficits,
which may be a consequence of altered axon branching
and/or growth. While it is not clear why a small number
of homozygous mutant mice are capable of surviving
into adulthood, it is possible that partial and variable reductions in the levels of nSR100 target AS events may afford sufficient degrees of diaphragm innervation in some
individuals. It is also known that inconsistent breathing
can be regularized by an entrainment contribution from
the respiratory center in the brainstem (Champagnat
et al. 2011; Giraudin et al. 2012; Mellen and Thoby-Brisson 2012). It is therefore possible that sufficient entrainment may occur early enough in a small fraction of the
homozygous mutants to allow for their survival. In any
case, to our knowledge, the deficit in phrenic nerve
branching in nSR100 mutants has not been described previously in mice deficient of other splicing regulators.
Another
neurodevelopmental
aberration
in
nSR100 78/78 mice that has not been previously
observed in other splicing factor knockouts is the axon
midline crossing defect in the corpus callosum. Approximately 60 mouse genes are known to be required for the
formation of the corpus callosum (Paul et al. 2007; Donahoo and Richards 2009). Interestingly, our RNA-seq analysis revealed that transcripts from one of these genes,
Slit2, contain one of the most strongly nSR100-dependent
exons. Slit2 is secreted by distinct cell populations located
at or near the midline. It binds Robo receptors expressed in
growing axons to help mediate midline crossing. Its function has been extensively studied in vivo (Chedotal 2007),
and a Slit2 knockout mouse displays a midline crossing
defect that is strikingly similar to the one that we observed in nSR100 78/78 mice, with bundles of callosal
axons projecting ventrally along the midline (Unni et al.
2012). The nSR100-dependent Slit2 microexon adds nine
amino acids to the fifth EGF domain in the secreted N-terminal portion of the protein that is responsible for its repulsive activity during axon guidance. The differential
activities of the resulting Slit2 isoforms have not been previously investigated, although an AS event in Robo3 that
switches the axonal response to Slit proteins from attraction to repulsion has been reported (Chen et al. 2008). It is
interesting to consider that the nSR100-dependent regulation of the alternative Slit2 exon represents a complementary mechanism for controlling axon guidance and
may contribute to the midline crossing defect observed
in nSR100 mutant mice. However, because the loss of

Downloaded from genesdev.cshlp.org on December 15, 2016 - Published by Cold Spring Harbor Laboratory Press

In vivo functions of nSR100

nSR100 also affects axon growth and the number and distribution of neurons that project callosal axons, we cannot
exclude the possibility that a combination of these mechanisms may contribute to this anomaly.
In addition to the differences observed in the corpus callosum of nSR100 heterozygotes and homozygotes, we discovered nSR100 dosage-dependent cortical deficits. It is
noteworthy that subtle deficits in the corpus callosum
and cortical layering have been linked to impaired cognitive and behavioral function in humans (Paul et al. 2007).
For example, disruption of cortical layer formation has
been observed in individuals with schizophrenia and autism (Akbarian et al. 1993; Ross et al. 2006; Stoner et al.
2014). While cell migration defects often result in the aberrant positioning of cortical layers (Caviness 1982; Kwan
et al. 2008), early production of post-mitotic neurons
by cortical progenitors has been shown to result in an expansion of deep cortical layers (Feng and Walsh 2004).
Premature production of neurons depletes the pool of progenitors and results in fewer late-born neurons being generated. In nSR100 78/78 mice, we found that layer VI is
significantly expanded in mutant brains at E18.5, whereas
the total number of neurons (including the number of
superficial neurons) and neural progenitors is decreased.
These phenotypes suggest that loss of nSR100 causes neural progenitors to fail to accomplish asymmetric division
early during brain development and commit prematurely
to a post-mitotic fate. Further supporting this proposal is
our finding that more neurons are produced during the
first steps of cortical development when nSR100 is lost.
Finally, although the preplate was poorly defined in the
brains of mutant mice, our fate-mapping experiments provide evidence that early-born, nSR100-depleted neurons
migrate correctly to populate deep cortical layers. These
observations thus establish nSR100 as an important regulator of neurogenesis timing.
nSR100 regulates AS events in genes with important
neuronal functions
Our previous studies directed at the identification of
nSR100 targets in cell lines focused on cassette alternative exons and revealed that nSR100 predominantly promotes neural exon inclusion (Calarco et al. 2009; Raj
et al. 2011; Irimia et al. 2014). The RNA-seq analyses performed here confirmed this property of nSR100 and further encompassed a comprehensive survey of all classes
of nSR100-regulated AS events detected in vivo. Notably,
65 of the 138 (47%) (longer) cassette exons and microexons that are promoted by nSR100 in vivo had not previously been reported as nSR100 targets. Many of these events
occur in genes with essential roles in neuronal biology
and/or nervous system development (e.g., Ahi1, Camk2b,
Dvl1, Itsn1, Shank1, and Unc13b). Moreover, in addition
to changes in the inclusion levels of a large number of neural cassette exons, of which many are microexons (see below), the present work also revealed that the nSR100
regulatory program extends to all classes of AS events.
For example, many retained introns are misregulated
in developing nSR100 78/78 mouse brains. Although a

subset of the retained introns introduce premature termination codons, it appears that in most cases, the corresponding transcripts are not subject to nonsense-mediated
mRNA decay, as their steady-state levels were not appreciably affected in nSR100 78/78 brain tissue (data not
shown). We also identified a small number of nSR100-dependent alternative 5 and 3 splice site selection events,
most of which are frame-preserving.
Collectively, AS events misregulated in nSR10078/78
mouse brains are enriched in genes involved in neuronal
functions, such as genes associated with neuronal differentiation (Zmynd8 and Ahi1), neurite outgrowth (Zfyve27
and Clasp2), and axon guidance (Slit2, Nrcam, and
Mycbp2). Many of these genes possess pivotal roles as
scaffolding proteins for endocytosis, exocytosis, cytoskeleton remodeling, and vesicle transport and are associated
with defects similar to the ones that we observed in our
mouse model.
Functional impact of nSR100-regulated microexons
Among genes that contain microexons regulated by
nSR100, several encode proteins that are known to interact. These proteins form a network that is involved in
the trafficking and recycling of vesicles, including Itsn1,
Ppfia2, Rims2, Dnm2, Nbea, Abi1, Ptprd, and Vav2. Strikingly, 65 of the 72 nSR100-activated microexons are
frame-preserving and have the potential to result in the insertion of one to nine amino acid residues in the corresponding protein products. These seemingly modest
changes to coding sequence raise interesting questions
as to the functional roles of microexons.
Recently, we and others have observed that amino acid
residues encoded by microexons are almost invariably
surface-accessible and enriched withinor immediately
adjacent todomains involved in proteinprotein or proteinlipid interactions (Irimia et al. 2014; Li et al. 2015).
Consistent with these observations, deletion of microexons reduces interactions with partner proteins. For example, a microexon in the SH3 domain of the Down
syndrome-associated gene Itsn1, which we show here is
strongly regulated by nSR100 (Supplemental Fig. S5), promotes interactions with multiple partners (Tsyba et al.
2008). A recent report demonstrated that an nSR100-regulated microexon in the Zfyve27 transcript (Fig. 5E) promotes interactions with partner proteins VAP-A and
VAP-B (Ohnishi et al. 2014). Furthermore, we showed recently that neural microexons in the AP1 endocytic transport complex subunit Ap1s2 and the amyloid precursor
protein-binding family B member 1 (Apbb1), which is also
associated with neuritogenesis (Cheung et al. 2014), promote interactions with their respective partner proteins
(Irimia et al. 2014). In the present study, we extend these
findings by demonstrating that nSR100-dependent inclusion of a 6-nt microexon in Unc13b transcripts is sufficient to promote the increased length of neurites and
rescue a neuritogenesis defect in nSR100 mutant neurons.
This microexon has the potential to add two amino acids
adjacent to a predicted MAPK-docking site in the Unc13b
protein. It is therefore interesting to consider that the

GENES & DEVELOPMENT

755

Downloaded from genesdev.cshlp.org on December 15, 2016 - Published by Cold Spring Harbor Laboratory Press

Quesnel-Vallires et al.

nSR100-dependent regulation of this microexon affects

the phosphorylation status of Unc13b in ways that alter
interactions required for neurite formation.
Finally, it is intriguing to note that most (76%) of the
microexons affected by the in vivo loss of nSR100 in the
present study are conserved in humans, and many
(46%) of these display loss of inclusion in the brain cortices of subjects with ASD (Irimia et al. 2014). Furthermore, this altered pattern of inclusion in ASD subjects
affects genes enriched in known genetic associations
with ASD and is also highly correlated with reduced expression of nSR100 (Irimia et al. 2014). Additional studies
have linked microexon misregulation to schizophrenia
and epilepsy (Ovadia and Shifman 2011; Rusconi et al.
2014). In the future, it will be of considerable interest to
address whether partial alteration of nSR100 levels in developing mice may contribute to phenotypes associated
with human neurological disorders.

Materials and methods

nSR100 mutant mouse generation
Stem cells containing the conditional nSR100 lox allele were
obtained from EUCOMM (project no. 71507, clones
EPD0538_3_A08 and EPD0538_3_A09) (Friedel et al. 2007) and
aggregated with outbred ICR morula. Following confirmation of
germline transmission, mice bearing the nSR100 lox allele were
maintained on a C57Bl/6N background and crossed with the
B6.C-Tg(CMV-cre)1Cgn/J line from the Jackson Laboratory. Excision of exons 7 and 8 in the resulting nSR100 +/78 mice was confirmed by PCR and sequencing.
We attempted to generate mice lacking the partial (25-kDa)
polypeptide detected in the nSR100 78/78 strain (Fig. 1) by removing both the Frt site-flanked LacZ reporter and exons 7 and
8 by sequential Flpase and Cre recombinase crosses (Supplemental Fig. S9A). However, the resulting nSR100 GT-ex8 allele produced an nSR100 protein migrating at 90 kDa, which is
consistent with the predicted size of a partial nSR100 protein
lacking only exons 7 and 8 (Supplemental Fig. S9B). We therefore
focused our analyses on the nSR100 78 allele.

Southern blotting
Southern blotting was performed as described elsewhere (Sambrook and Russell 2001). Briefly, 60 g of mouse genomic tail
DNA was digested with AseI and loaded on a 0.8% agarose gel
for each genotype. DNA was transferred to a Hybond XL membrane (GE Healthcare Life Sciences) and hybridized with a
32
P-dCTP-labeled probe encompassing 456 base pairs (bp) of
nSR100 intron 3 upstream of the 5 homology arm used for homologous recombination of the nSR100 lox allele.

RTPCR
Semi-quantitative RTPCR was performed using the Qiagen onestep RTPCR kit according to the manufacturers instructions using 15 ng of total RNA template per 10-L reaction and run on 2%
or 4% agarose gels. Radiolabeled reactions included 0.05 Ci of
32
P-dCTP and were run on 6% Sequagel urea gels (National Diagnostics). Bands were quantified using Image Lab (Bio-Rad) or
ImageJ. Primer sequences are available on request.

756

GENES & DEVELOPMENT

Antibodies
For immunoblotting, a polyclonal rabbit antibody (Calarco et al.
2009) raised against amino acids 182 of nSR100 was used at
1:5000. Anti-tubulin (T6074, Sigma) was used at 1:5000. For
immunostaining, mouse monoclonal anti-neurofilament (2H3
conditioned medium, Iowa Developmental Studies Hybridoma
Bank) was diluted to 1:50 for whole-mount diaphragm staining
and 1:100 for brain section staining. Mouse anti-NeuN
(mab377, Millipore), mouse anti-Satb2 (ab51502, Abcam), rabbit
anti-Tbr1 (ab31940, Abcam), and chicken polyclonal anti--galactosidase (ab9361, Abcam) were all diluted to 1:500. Chicken antiMAP2 (ab5392, Abcam) was diluted to 1:10,000, mouse anti-Tuj1
(MRB-435P, Covance) was diluted to 1:750, and rabbit anti-Pax6
(PRB-278P, Covance) was diluted to 1:1500. For in situ hybridization, an anti-DIG antibody conjugated to alkaline phosphatase
(Roche) was diluted to 1:5000.
In situ hybridization
In situ hybridization was essentially performed as previously described (Sambrook and Russell 2001). Twenty-micrometer brain
sections were post-fixed in 4% formaldehyde for 10 min at
room temperature. Sections were then prehybridized for 36 h
at room temperature followed by hybridization with sense or antisense DIG-labeled probes to nSR100 exons 913 diluted to 200
ng/mL overnight at 60C. Alkaline-phosphatase-conjugated
anti-DIG antibody was added to slides for 1 h at room temperature
and washed, and an NBT/BCIP solution (Roche) was applied for
1 h to overnight at room temperature. Sections were cleared in xylene and mounted in Cytoseal XYL (Thermo Scientific).
Immunofluorescence
For whole-mount diaphragm staining, diaphragms were dissected
from E16.5 or E18.5 embryos and fixed in 2% formaldehyde overnight at 4C. Diaphragms were washed in 0.1 M glycine in PBS
and blocked overnight at 4C in 0.5% Triton X-100, 3% BSA,
and 5% donkey serum with Alexa-594-coupled -bungarotoxin
diluted at 1:1500 (Life Technologies). Diaphragms were then further permabilized briefly in 100% methanol, fixed again in 0.2%
glutaraldehyde and 4% formaldehyde for 20 min at room temperature, and then incubated overnight at 4C in blocking buffer
with a monoclonal anti-neurofilament antibody diluted to
1:100. After extensive washes, samples were incubated overnight
at 4C in blocking buffer with an Alexa-488 anti-mouse antibody
diluted to 1:500 (Life Technologies).
For immunofluorescence microscopy of cortical sections, embryonic mouse brains were dissected and fixed in 4% formaldehyde for 60 min at 4C. Brains were then sucrose-protected,
OCT-embedded (Tissue-Tek), frozen, and coronally sectioned to
16 m. Sections were blocked and permeabilized in 5% BSA
and 0.3% Triton X-100 in PBS with or without M.O.M. blocking
reagent (Vector Laboratories) for 60 min at room temperature. All
primary antibodies were incubated overnight at 4C with or without M.O.M. protein concentrate. Secondary antibodies were goat
anti-mouse, anti-rabbit, or anti-chicken coupled to Alexa-488,
Alexa-568, or Alexa-594 dyes, all diluted to 1:500 and incubated
for 60 min at room temperature with or without M.O.M. protein
concentrate. Stained sections were mounted in VectaShield
mounting medium with DAPI (Vector Laboratories).
EdU labeling
Pregnant dams were injected with 0.1 mg/g EdU at E12.5. Embryos were harvested at E18.5, and brains were dissected, fixed, and

Downloaded from genesdev.cshlp.org on December 15, 2016 - Published by Cold Spring Harbor Laboratory Press

In vivo functions of nSR100

cryosectioned as described above. Immunostaining using an

antibody to Tbr1 was first performed as described above. EdU+
cells were then stained using the Click-IT Alexa Fluor-488 kit
(Life Technologies) according to the manufacturers protocol.
Quantification of phenotypic data
For neurite length measurements on whole-mount diaphragms,
tracings were generated with the NeuronJ plug-in for ImageJ.
Right branches on the ventral and dorsal parts of the diaphragm
were measured and counted. For NMJs, a 475-m-long region
was selected over the ventrally projecting left primary branch,
and NMJs were quantified using the ITCN plug-in in ImageJ.
The dispersion of NMJs was measured as the average width of
the NMJ band at three different levels of the same region of interest. Numbers of cells were counted using the ITCN plug-in over
300-m radial unit regions for Tbr1+, Satb2+, NeuN+, and Pax6+
cells and a 750-m radial unit region for EdU+ cells. Cells and layers were quantified using three sections from each brain. SVZ
thickness was measured relative to the total thickness of the cerebral cortex from the lateral ventricle to the surface of layer I.
RNA-seq analysis
A first replicate consisted of total RNA extracted from cerebral
cortices and hippocampi dissected from five wild-type and five
nSR100 78/78 mutant brains at E18.5. RNA was pooled by genotype and prepared using the Illumina TruSeq mRNA kit, and
cortical and hippocampal samples were sequenced on different
runs of Illumina HiSeq2500 (average of 93 million 100-nt single
end and 100-nt paired-end reads for each run, respectively) (see
Supplemental Table S4 for details). A second replicate was processed as above and consisted of total RNA pooled from three
wild-type or three mutant brains at E18.5. An average of 90 million 100-nt paired-end reads was sequenced for each sample (Supplemental Table S4).
Transcriptome-wide AS and gene expression profiling was performed using our recently described pipeline (vast-tools) (Irimia
et al. 2014). vast-tools uses reads mapping to exonexon (or
exonintron) junctions (EEJ or EIJ) only to accurately detect and
quantify all types of AS events, including 3- to 27-nt microexons.
Gene expression levels were measured using the cRPKM metric
(Labbe et al. 2012).
PSI/PIR/PSU of AS events for the eight samples were paired
into four replicates (wild-type and nSR100 78/78 for two cortex
and two hippocampus samples), and a paired t-test was performed
for AS events with enough read coverage in all eight samples. A
given AS event was considered to have sufficient read coverage
in a particular RNA-seq sample according to the following criteria (Irimia et al. 2014):
For cassette exons (except for those quantified using the microexon pipeline), (1) 10 actual reads (i.e., before mappability correction) mapped to the sum of exclusion EEJs or (2) 10 actual
reads mapped to one of the two inclusion EEJs and five or more
mapped to the other inclusion EEJ.
For microexons, (1) 10 actual reads mapped to the sum of exclusion EEJs or (2) 10 actual reads mapped to the sum of inclusion EEJs.
For IR, (1) 10 actual reads mapped to the sum of skipping EEJs
or (2) 10 actual reads mapped to one of the two inclusion EIJs
and five or more mapped to the other inclusion EIJ.
For Alt3 and Alt5, 10 actual reads mapped to the sum of all
EEJs involved in the specific event.

For an AS event to be considered differentially regulated between

wild-type and nSR100 78/78 brains, we required a P-value of
<0.05 in the t-test and an average PSI (between the four paired
replicates) of at least 10%.
All RNA-seq data have been deposited in the Gene Expression
Omnibus database under accession number GSE65818.

Functional enrichment analyses

Ensembl gene IDs for the cassette exons and microexons that
showed significantly increased skipping in nSR100 78/78 brains
(137 genes in total) were uploaded to DAVID (http://david.abcc.
ncifcrf.gov; Huang et al. 2009) to perform functional enrichment
analyses using a stringent background consisting of 10,968 genes
with expression of at least cRPKM >2 in one of the brain samples.
Only GO_FAT terms and KEGG (Kyoto Encyclopedia of Genes
and Genomes) pathways were used for clustering analyses.

Primary neuronal cultures

Protocols for culturing primary mouse neurons were kindly provided by Dr. Antony Boucard and Dr. Thomas Sudhof (Stanford
University) (Boucard et al. 2005). Briefly, hippocampal or cortical
neurons were harvested from wild-type or nSR100 78/78 mice
at E18.5 and plated on glass coverslips coated with 2% Matrigel
(Corning) in plating medium consisting of MEM (51200-038,
Life Technologies) supplemented with 0.5% glucose, 0.2 mg/
mL NaHCO3, 0.1 mg/mL transferrin (616420, Calbiochem),
10% fetal bovine serum (FBS) (SH30396.03, GE Life Sciences),
2 mM L-glutamine (12403-010, Life Technologies), and 25 g/
mL insulin (I-6634, Sigma). Plating medium was changed at
DIV1 to growth medium consisting of MEM supplemented
with 0.5% glucose, 0.2 mg/mL NaHCO3, 0.1 mg/mL transferrin,
5% FBS, 0.5 mM L-glutamine, and 2% B-27 supplement (17504044, Life Technologies). Dissociated neurons were transfected
prior to plating using the Amaxa Nucleofector kit (VPG-1001,
Lonza) using 5 104 cells and 10 g of plasmid DNA per
transfection. Cells from the same embryos were aliquoted and
transfected individually with each construct. Unc13_skp,
Unc13_inc, and nSR100 were cloned upstream of the RFP coding
sequence and placed under the control of the CAGGS promoter.
The length of neurites was quantified using the NeuronJ plug-in
for ImageJ.

Acknowledgments
We thank Marina Gertsenstein, Sandra Tondat, Monica Pereira,
Christina Dalrymple, Jorge Cabezas, Jessica Raponi, and Amanda
Leonelli at the Toronto Centre for Phenogenomics for assistance
with the generation and maintenance of mutant mouse strains.
Dax Torti and Danica Leung of the Donnelly Sequencing Centre
are gratefully acknowledged for sequencing samples. We also
thank Jonathan Ellis for generating constructs used in primary
neuronal cultures, and members of the Blencowe and Cordes laboratories for helpful discussions and comments on the manuscript. This research was supported by grants from the
Canadian Institutes of Health Research (CIHR; MOP-67011,
MOP-14609, and MOP-111199) to B.J.B. and S.P.C. M.I. was supported by a long-term Fellowship from the Human Frontiers Science Program Organization. M.Q.-V. was supported by CIHR
Banting and Best Scholarship and Ontario Graduate Scholarship.
B.J.B. holds the Banbury Chair of Medical Research at the University of Toronto.

GENES & DEVELOPMENT

757

Downloaded from genesdev.cshlp.org on December 15, 2016 - Published by Cold Spring Harbor Laboratory Press

Quesnel-Vallires et al.

References
Akbarian S, Bunney WE Jr, Potkin SG, Wigal SB, Hagman JO,
Sandman CA, Jones EG. 1993. Altered distribution of nicotinamide-adenine dinucleotide phosphate-diaphorase cells in
frontal lobe of schizophrenics implies disturbances of cortical
development. Arch Gen Psychiatry 50: 169177.
Barbosa-Morais NL, Irimia M, Pan Q, Xiong HY, Gueroussov S,
Lee LJ, Slobodeniuc V, Kutter C, Watt S, Colak R, et al.
2012. The evolutionary landscape of alternative splicing in
vertebrate species. Science 338: 15871593.
Boucard AA, Chubykin AA, Comoletti D, Taylor P, Sudhof TC.
2005. A splice code for trans-synaptic cell adhesion mediated
by binding of neuroligin 1 to - and -neurexins. Neuron 48:
229236.
Braunschweig U, Gueroussov S, Plocik AM, Graveley BR, Blencowe BJ. 2013. Dynamic integration of splicing within gene
regulatory pathways. Cell 152: 12521269.
Braunschweig U, Barbosa-Morais NL, Pan Q, Nachman EN, Alipanahi B, Gonatopoulos-Pournatzis T, Frey B, Irimia M, Blencowe BJ. 2014. Widespread intron retention in mammals
functionally tunes transcriptomes. Genome Res 24: 1774
1786.
Broeke JHP, Roelandse M, Luteijn MJ, Boiko T, Matus A, Toonen
RF, Verhage M. 2010. Munc18 and Munc13 regulate early
neurite outgrowth. Biol Cell 102: 479488.
Buckanovich RJ, Posner JB, Darnell RB. 1993. Nova, the paraneoplastic Ri antigen, is homologous to an RNA-binding protein
and is specifically expressed in the developing motor system.
Neuron 11: 657672.
Buljan M, Chalancon G, Eustermann S, Wagner GP, Fuxreiter M,
Bateman A, Babu MM. 2012. Tissue-specific splicing of disordered segments that embed binding motifs rewires protein interaction networks. Mol Cell 46: 871883.
Calarco JA, Superina S, OHanlon D, Gabut M, Raj B, Pan Q, Skalska U, Clarke L, Gelinas D, van der Kooy D, et al. 2009. Regulation of vertebrate nervous system alternative splicing and
development by an SR-related protein. Cell 138: 898910.
Caviness VSJ. 1982. Neocortical histogenesis in normal and reeler mice: a developmental study based upon [3H]thymidine autoradiography. Brain Res 256: 293302.
Champagnat J, Morin-Surun MP, Bouvier J, Thoby-Brisson M,
Fortin G. 2011. Prenatal development of central rhythm generation. Respir Physiol Neurobiol 178: 146155.
Chedotal A. 2007. Slits and their receptors. Adv Exp Med Biol
621: 6580.
Chen M, Manley JL. 2009. Mechanisms of alternative splicing
regulation: insights from molecular and genomics approaches.
Nat Rev Mol Cell Biol 10: 741754.
Chen Z, Gore BB, Long H, Ma L, Tessier-Lavigne M. 2008. Alternative splicing of the Robo3 axon guidance receptor governs
the midline switch from attraction to repulsion. Neuron 58:
325332.
Cheung HN, Dunbar C, Morotz GM, Cheng WH, Chan HY, Miller CC, Lau KF. 2014. FE65 interacts with ADP-ribosylation
factor 6 to promote neurite outgrowth. FASEB J 28: 337349.
Deol MS, Gluecksohn-Waelsch S. 1979. The role of inner hair
cells in hearing. Nature 278: 250252.
Donahoo AL, Richards LJ. 2009. Understanding the mechanisms
of callosal development through the use of transgenic mouse
models. Semin Pediatr Neurol 16: 127142.
Ellis JD, Barrios-Rodiles M, Colak R, Irimia M, Kim T, Calarco JA,
Wang X, Pan Q, OHanlon D, Kim PM, et al. 2012. Tissue-specific alternative splicing remodels proteinprotein interaction
networks. Mol Cell 46: 884892.

758

GENES & DEVELOPMENT

Feng Y, Walsh CA. 2004. Mitotic spindle regulation by Nde1 controls cerebral cortical size. Neuron 44: 279293.
Friedel RH, Seisenberger C, Kaloff C, Wurst W. 2007. EUCOMM
the European conditional mouse mutagenesis program.
Brief Funct Genomic Proteomic 6: 180185.
Gao P, Sultan KT, Zhang XJ, Shi SH. 2013. Lineage-dependent circuit assembly in the neocortex. Development 140:
26452655.
Gehman LT, Meera P, Stoilov P, Shiue L, OBrien JE, Meisler MH,
Ares MJr, Otis TS, Black DL. 2012. The splicing regulator
Rbfox2 is required for both cerebellar development and mature motor function. Genes Dev 26: 445460.
Giraudin A, Le Bon-Jego M, Cabirol MJ, Simmers J, Morin D.
2012. Spinal and pontine relay pathways mediating respiratory rhythm entrainment by limb proprioceptive inputs in the
neonatal rat. J Neurosci 32: 1184111853.
Huang DW, Sherman BT, Lempicki RA. 2009. Bioinformatics enrichment tools: paths toward the comprehensive functional
analysis of large gene lists. Nucleic Acids Res 37: 113.
Irimia M, Weatheritt RJ, Ellis JD, Parikshak NN, GonatopoulosPournatzis T, Babor M, Quesnel-Vallieres M, Tapial J, Raj B,
OHanlon D, et al. 2014. A highly conserved program of neuronal microexons is misregulated in autistic brains. Cell
159: 15111523.
Kwan KY, Lam MM, Krsnik Z, Kawasawa YI, Lefebvre V, Sestan
N. 2008. SOX5 postmitotically regulates migration, postmigratory differentiation, and projections of subplate and deeplayer neocortical neurons. Proc Natl Acad Sci 105: 16021
16026.
Labbe RM, Irimia M, Currie KW, Lin A, Zhu SJ, Brown DD, Ross
EJ, Voisin V, Bader GD, Blencowe BJ, et al. 2012. A comparative transcriptomic analysis reveals conserved features of
stem cell pluripotency in planarians and mammals. Stem
Cells 30: 17341745.
Li Q, Lee JA, Black DL. 2007. Neuronal regulation of alternative
pre-mRNA splicing. Nat Rev Neurosci 8: 819831.
Li Q, Zheng S, Han A, Lin CH, Stoilov P, Fu XD, Black DL. 2014.
The splicing regulator PTBP2 controls a program of embryonic
splicing required for neuronal maturation. eLife 3: e01201.
Li YI, Sanchez-Pulido L, Haerty W, Ponting CP. 2015. RBFOX and
PTBP1 proteins regulate the alternative splicing of micro-exons in human brain transcripts. Genome Res 25: 113.
Licatalosi DD, Yano M, Fak JJ, Mele A, Grabinski SE, Zhang C,
Darnell RB. 2012. Ptbp2 represses adult-specific splicing to
regulate the generation of neuronal precursors in the embryonic brain. Genes Dev 26: 16261642.
Lipscombe D. 2005. Neuronal proteins custom designed by alternative splicing. Curr Opin Neurobiol 15: 358363.
Lipscombe D, Allen SE, Toro CP. 2013a. Control of neuronal voltage-gated calcium ion channels from RNA to protein. Trends
Neurosci 36: 598609.
Lipscombe D, Andrade A, Allen SE. 2013b. Alternative splicing:
functional diversity among voltage-gated calcium channels
and behavioral consequences. Biochim Biophys Acta 1828:
15221529.
Mellen NM, Thoby-Brisson M. 2012. Respiratory circuits: development, function and models. Curr Opin Neurobiol 22:
676685.
Merkin J, Russell C, Chen P, Burge CB. 2012. Evolutionary dynamics of gene and isoform regulation in mammalian tissues.
Science 338: 15931599.
Nakano Y, Jahan I, Bonde G, Sun X, Hildebrand MS, Engelhardt
JF, Smith RJ, Cornell RA, Fritzsch B, Banfi B. 2012. A mutation
in the Srrm4 gene causes alternative splicing defects and deafness in the Bronx waltzer mouse. PLoS Genet 8: e1002966.

Downloaded from genesdev.cshlp.org on December 15, 2016 - Published by Cold Spring Harbor Laboratory Press

In vivo functions of nSR100

Norris AD, Calarco JA. 2012. Emerging roles of alternative premRNA splicing regulation in neuronal development and function. Front Neurosci 6: 122.
Ohnishi T, Shirane M, Hashimoto Y, Saita S, Nakayama KI. 2014.
Identification and characterization of a neuron-specific isoform of protrudin. Genes Cells 19: 97111.
Ovadia G, Shifman S. 2011. The genetic variation of RELN expression in schizophrenia and bipolar disorder. PLoS One 6:
e19955.
Paul LK, Brown WS, Adolphs R, Tyszka JM, Richards LJ, Mukherjee P, Sherr EH. 2007. Agenesis of the corpus callosum: genetic, developmental and functional aspects of connectivity. Nat
Rev Neurosci 8: 287299.
Raj B, OHanlon D, Vessey JP, Pan Q, Ray D, Buckley NJ, Miller
FD, Blencowe BJ. 2011. Cross-regulation between an alternative splicing activator and a transcription repressor controls
neurogenesis. Mol Cell 43: 843850.
Raj B, Irimia M, Braunschweig U, Sterne-Weiler T, OHanlon D,
Lin ZY, Chen GI, Easton LE, Ule J, Gingras AC, et al. 2014.
A global regulatory mechanism for activating an exon network required for neurogenesis. Mol Cell 56: 90103.
Ross CA, Margolis RL, Reading SA, Pletnikov M, Coyle JT. 2006.
Neurobiology of schizophrenia. Neuron 52: 139153.
Ruggiu M, Herbst R, Kim N, Jevsek M, Fak JJ, Mann MA, Fischbach G, Burden SJ, Darnell RB. 2009. Rescuing Z+ agrin splicing in Nova null mice restores synapse formation and
unmasks a physiologic defect in motor neuron firing. Proc
Natl Acad Sci 106: 35133518.
Rusconi F, Paganini L, Braida D, Ponzoni L, Toffolo E, Maroli A,
Landsberger N, Bedogni F, Turco E, Pattini L, et al. 2014.
LSD1 neurospecific alternative splicing controls neuronal excitability in mouse models of epilepsy. Cereb Cortex doi:
10.1093/cercor/bhu070.

Sambrook J, Russell DW. 2001. Molecular cloning: a laboratory

manual. Cold Spring Harbor Laboratory Press, Cold Spring
Harbor, NY.
Shen H, Green MR. 2004. A pathway of sequential arginineserine-rich domain-splicing signal interactions during mammalian spliceosome assembly. Mol Cell 16: 363373.
Stoner R, Chow ML, Boyle MP, Sunkin SM, Mouton PR, Roy S,
Wynshaw-Boris A, Colamarino SA, Lein ES, Courchesne E.
2014. Patches of disorganization in the neocortex of children
with autism. N Engl J Med 370: 12091219.
Tsyba L, Gryaznova T, Dergai O, Dergai M, Skrypkina I, Kropyvko S, Boldyryev O, Nikolaienko O, Novokhatska O, Rynditch A. 2008. Alternative splicing affecting the SH3A
domain controls the binding properties of intersectin 1 in neurons. Biochem Biophys Res Commun 372: 929934.
Ule J, Darnell RB. 2006. RNA binding proteins and the regulation of
neuronal synaptic plasticity. Curr Opin Neurobiol 16: 102110.
Unni DK, Piper M, Moldrich RX, Gobius I, Liu S, Fothergill T,
Donahoo AL, Baisden JM, Cooper HM, Richards LJ. 2012. Multiple Slits regulate the development of midline glial populations and the corpus callosum. Dev Biol (N Y 1985) 365: 3649.
Wu JY, Maniatis T. 1993. Specific interactions between proteins
implicated in splice site selection and regulated alternative
splicing. Cell 75: 10611070.
Yang YY, Yin GL, Darnell RB. 1998. The neuronal RNA-binding
protein Nova-2 is implicated as the autoantigen targeted in
POMA patients with dementia. Proc Natl Acad Sci 95:
1325413259.
Yano M, Hayakawa-Yano Y, Mele A, Darnell RB. 2010. Nova2
regulates neuronal migration through an RNA switch in disabled-1 signaling. Neuron 66: 848858.
Zheng S, Black DL. 2013. Alternative pre-mRNA splicing in neurons: growing up and extending its reach. Trends Genet 29:
442448.

GENES & DEVELOPMENT

759

Downloaded from genesdev.cshlp.org on December 15, 2016 - Published by Cold Spring Harbor Laboratory Press

Essential roles for the splicing regulator nSR100/SRRM4 during

nervous system development
Mathieu Quesnel-Vallires, Manuel Irimia, Sabine P. Cordes, et al.
Genes Dev. 2015 29: 746-759
Access the most recent version at doi:10.1101/gad.256115.114

Supplemental
Material
References

http://genesdev.cshlp.org/content/suppl/2015/04/01/29.7.746.DC1.html

This article cites 55 articles, 15 of which can be accessed free at:

http://genesdev.cshlp.org/content/29/7/746.full.html#ref-list-1

Creative
Commons
License

This article is distributed exclusively by Cold Spring Harbor Laboratory Press for the first
six months after the full-issue publication date (see
http://genesdev.cshlp.org/site/misc/terms.xhtml). After six months, it is available under a
Creative Commons License (Attribution-NonCommercial 4.0 International), as described
at http://creativecommons.org/licenses/by-nc/4.0/.

Email Alerting
Service

Receive free email alerts when new articles cite this article - sign up in the box at the top
right corner of the article or click here.

To subscribe to Genes & Development go to:

http://genesdev.cshlp.org/subscriptions

2015 Quesnel-Vallires et al.; Published by Cold Spring Harbor Laboratory Press